rite ghg mitigation assessment model dne21+...2015/01/30 · january 30, 2015 systems analysis...
TRANSCRIPT
1
Research Institute of Innovative
Technology for the Earth
RITE GHG Mitigation Assessment Model DNE21+
January 30, 2015
Systems Analysis Group
Research Institute of Innovative Technology for the Earth (RITE)
1. Model scope and methods DNE21+
1.1. Model concept, solver and details
RITE’s integrated assessment framework consists of 3 modules; 1) Key Assessment
Model DNE21+ for energy-related CO2, 2) Non-energy CO2 emission scenario, which
assumes specific non-energy CO2 emissions separately from mitigation levels of
energy-related CO2 emissions 3) Non-CO2 GHG Assessment Model, for mitigation of the
five non-CO2 greenhouse gases emissions of the Kyoto Protocol, based on the United
States Environmental Protection Agency (US EPA) assessments.
Figure 1 Integrated Assessment Framework
DNE21+ is an intertemporal linear programming model for assessing global energy
systems and global warming mitigation options. It represents energy systems (e.g.,
2
Research Institute of Innovative
Technology for the Earth
energy flows, capacities of energy related facilities) consistently in which the sum of the
discounted total energy systems costs are minimized. The model represents regional
differences and assesses detailed energy-related CO2 emission reduction technologies
up to 2050. The energy supply sectors are hard-linked with energy end-use sectors,
including energy exporting/importing; working lifetimes of facilities are taken into
account so that assessments maintain complete consistency over energy systems and
over time periods.
The base year of the model is 2000, and the GHG emissions are completely consistent
with the historical data. The historical data on total GHG emissions for Annex I and
non-Annex I parties are based on the GHG inventories of the UNFCCC and IEA
statistics respectively. Energy-related CO2 emissions are based on the IEA statistics for
all the countries. Whereas the statistical data for energy-related CO2 emissions differ
between the UNFCCC and IEA in some countries, non-CO2 GHG emissions for Annex I
parties are defined by subtracting the energy-related CO2 emissions reported by the
IEA and the non-energy use CO2 emissions inventory of the UNFCCC from the total
GHG emissions of the UNFCCC, thus, giving priority to the total GHG emission
consistency with the UNFCCC inventory.
Technological costs and energy efficiency of technologies (various type of power
generation technologies, oil refinery, coal gasification technology, etc.) and carbon
dioxide capture, storage and sequestration are explicitly modeled. Energy intensive
industries such as steel, cement, paper & pulp, aluminum, some groups of the chemical
industry (ethylene, propylene production in the petrochemical industry and ammonia
production), transportation (automobiles) and several groups of residential &
commercial sector are also explicitly modeled ("Bottom-up approach”). The amounts of
activities of these sectors (industry: outputs, automobiles: transportation demands,
groups of residential & commercial sector: time periods of equipment utilization) are
estimated exogenously and kept fixed in this model regardless of emissions constraints,
while technological options are endogenously determined in the model. Other sectors,
whose technological characteristics and future evolutions vary widely, are modeled in a
top-down fashion.
(Reference)
Akimoto, Keigo, et al. "Comparison of marginal abatement cost curves for 2020 and
2030: longer perspectives for effective global GHG emission reductions." Sustainability
3
Research Institute of Innovative
Technology for the Earth
Science 7.2 (2012): 157-168.
Akimoto, Keigo, et al. "Estimates of GHG emission reduction potential by country, sector,
and cost." Energy Policy 38.7 (2010): 3384-3393.
1.2. Spatial process
DNE21+ has global coverage and divides the world into 54 regions (America, Canada,
Australia, China, India, Russia are divided into further small regions, making a total of
77 regions).
Figure 2 DNE21+ 77 regions
Table 1 List of DNE21+ regions
DNE21+ Region Country
United States United States, United States Virgin Islands
Guam, Puerto Rico
Canada Canada
United Kingdom United Kingdom
France France, Monaco
Germany Germany
Italy Italy, San Marino, Vatican City
Spain, Portugal Spain, Portugal, Azores (Port.)
Belgium, Netherlands, Denmark Belgium, Netherlands, Denmark
North Europe Sweden, Finland
Other EU Austria, Ireland, Greece, Luxembourg
4
Research Institute of Innovative
Technology for the Earth
Norway, Iceland Norway, Iceland
Greenland (Denmark) Greenland
Other Western Europe Switzerland, Liechtenstein, Malta, Andorra
Faeroe Islands, Gibraltar, Cyprus
Japan Japan
Australia Australia
New Zealand New Zealand
Other Oceania Papua New Guinea, Fiji, French Polynesia,
Kiribati, Nauru, New Caledonia, Solomon
Islands, Tonga, American Samoa, Vanuatu
China China, Hong Kong
North Korea, Mongolia Democratic People's Republic of Korea,
Mongolia
Viet Nam, Cambodia, Laos Viet Nam, Cambodia, Lao People's
Democratic Republic
Korea Korea
Malaysia, Singapore Malaysia, Singapore
Indonesia Indonesia, East Timor
Thailand Thailand
Philippines Philippines
Brunei Brunei Darussalam
Chinese Taipei Taiwan Province of China
India India
Pakistan, Afghanistan Pakistan, Afghanistan
Myanmar Myanmar
Other Asia Bangladesh, Nepal, Bhutan, Sri Lanka,
Maldives
Iran Iran
Saudi Arabia Saudi Arabia
Bahrain, Oman, Qatar, UAE, Yemen Bahrain, Oman, Qatar, United Arab
Emirates
Yemen
Other Middle East Iraq, Kuwait, Jordan, Israel, Lebanon,
Syrian Arab Republic
Turky Turkey
5
Research Institute of Innovative
Technology for the Earth
North Africa Egypt, Libyan Arab Jamahiriya, Tunisia,
Algeria, Morocco
South Africa South Africa
South East Africa Sudan, Eritrea, Djibouti, Ethiopia, Somalia,
Kenya, Uganda, Rwanda, Burundi, United
Republic of Tanzania, Malawi, Mozambique,
Swaziland, Lesotho, Madagascar, Seychelles
Comoros, Mauritius, Reunion
Other S.S.Africa Angola, Benin, Botswana, Burkina Faso,
Cameroon, Cape Verde, Central African
Republic, Chad, Congo, Cote d'Ivoire,
Democratic Republic of the Congo,
Equatorial Guinea, Gabon, Gambia, Ghana,
Guinea, Guinea-Bissau, Liberia, Mali,
Mauritania, Namibia, Niger, Nigeria, Sao
Tome and Principe, Senegal, Sierra Leone,
Togo, Zaire, Zambia, Zimbabwe, Western
Sahara
Mexico Mexico
Other Central America Bahamas, Bermuda, Cuba, Jamaica, Haiti,
El Salvador, Guadeloupe, Saint Vincent and
the Grenadines, Grenada, Dominica,
Dominican Republic, Saint Lucia, Saint Kitts
and Nevis, Barbados, Antigua & Barbuda,
Netherlands Antilles, Trinidad and Tobago,
Guatemala, Belize, Honduras, Nicaragua,
Costa Rica, Panama
Brazil Brazil
Venezuela, Guyana, Suriname Venezuela, Guyana, Suriname, French
Guiana
Paraguay, Uruguay, Argentina Paraguay, Uruguay, Argentina
Other South America Colombia, Ecuador, Peru, Bolivia, Chile
Russia Russian Federation
Other Annex I of FUSSR Ukraine, Estonia, Latvia, Lithuania
Belarus Belarus
Kazakhstan Kazakhstan
6
Research Institute of Innovative
Technology for the Earth
Other FUSSR Kyrgyzstan, Tajikistan, Turkmenistan,
Uzbekistan, Armenia, Azerbaijan, Georgia
OECD E.Europe Hungary, Poland, Czech Republic
Other Annex I of East Europe Bulgaria, Romania, Slovakia, Croatia,
Slovenia
Other E.Europe Yugoslavia, Albania, Bosnia And
Herzegovina
Republic Of Moldova, The former Yugoslav
Republic of Macedonia
1.3. Temporal process
DNE21+ is an inter-temporal optimization perfect foresight model with the time
horizon 2005 to 2050 in 5- and 10 year time steps where the first 6 periods (2005, 2010,
2015, 2020, 2025 and 2030) are 5-year periods and the remaining 2 periods are 10-year
periods(2040 and 2050). 2005 represents the period from 2003 to 2007, 2010 represents
the period from 2008 to 2012, 2015 represents the period from 2013 to 2017 and so on.
1.4. Policy
Based on the scenario setup, energy- and climate-related policies are explicitly
represented in DNE21+. This includes carbon pricing, emission cap and trade system,
carbon tax, preferred tax on specific energy sources, fuel subsidies, fuel standards and
energy standards. In general, these policies are implemented via constraints or a price
mark-up on energy sources.
When any emission restrictions (e.g., upper limit of emissions, emissions reduction
targets, specific unit improvement goals, carbon taxes) are applied, the model finds out
the energy systems whose costs are minimized, meeting all the assumed requirements,
given the sectoral amounts of production activities (e.g., crude steel and cement), the
amount of service activities (e.g., the traffic amount in the transportation sector), the
final energy demand in other sectors and the performances and the facility costs of
various technologies.
7
Research Institute of Innovative
Technology for the Earth
2. Economy and demand drivers
2.1. Population and GDP
The primary drivers of future energy demand in DNE21+ are projections of population
and GDP at market exchange rate (MER). We developed two different prospects as they
may more or less likely take place in the future as macroeconomic trends, considering
from the past statistics. In Scenario A, slow economic growth slows down after the
miraculous past growth mainly in developed countries In Scenario B, technological
advances keep ongoing as in the past and per capita GDP continues to grow quite
rapidly. As a reference case, we usually refer to Scenario A, but it depends on scope and
purpose of analysis.
The world population scenarios were developed with reference to The UN Population
Division; World Population Prospects (The 2008 Revision) which have been used
worldwide. UN scenarios of the world population are developed every two years and
have been revised downward for every update. Therefore, in this scenario, even after
taking account of the future population increase in developing countries such as in Asia
or Africa, we assume it very unlikely that the future world population will be
substantially over 10 billion. Historical statistics explicitly show the trends that the
fertility and population growth rates become lower with growing GDP per capita. Our
population scenario is developed, assuming this trend to keep in the future, by replacing
the relationship between fertility and per capita GDP by the relationship between
annual change rate of population and per capita GDP. The population growth in
Scenario B is assumed to be smaller than that in Scenario A, as per capita GDP is larger
in Scenario B.
Figure 3 shows the world population scenarios. In Scenario A, the world population is
assumed to have a medium growth rate and the UN medium variant scenario of the
world population, the 2008 Revision is adopted. After growing to 9.1 billion in 2050, the
world population grows steadily to 9.3 billion by the year 2100. In Scenario B, the world
population is assumed to have a low growth rate. This scenario is roughly equivalent to
the average of UN medium variant and low variant scenarios of the world population,
the 2008 Revision. The world population grows slowly from 6.1 billion in 2000. After
peaking at 8.6 billion around 2050, it declines to 7.4 billion by the year 2100.
8
Research Institute of Innovative
Technology for the Earth
Figure 3 Population Scenario
In terms of GDP per capita, the following distinct trends are observed historically.
The growth rate of GDP per capita is low in the least developed countries (LDCs).
When GDP per capita is between a few hundred dollars and a thousand dollars, the
growth rate of GDP per capita tends to become high.
For the higher GDP per capita, the growth rate tends to decrease gradually, shifting
toward moderate economic growth.
The industrial structure has three big trends; in the first period the structure
centers in primary industry, in high economic growth period heavy industry
develops starting from light industry, and in gradual growth period the tertiary
industry starts to grow such as service and information industries.
Based on these observations, GDP (MER) per capita scenarios are developed as shown
Figure 4. In Scenario A, the current developed countries slow down the GDP per capita
growth until 2100 and the growth rate converges to 0.5% per year in 2100. Developing
countries continue to grow steadily. The current emerging economies and least
developed countries have the per capita GDP growth rates of around 1%/year and
around 2%/year in 2100, respectively. The global average growth rate from 2000 to 2100
is 1.5% per year.
In Scenario B, the current developed countries continue to increase GDP per capita by
1.0%/year in 2100. Developing countries continue to grow rapidly. The current emerging
9
Research Institute of Innovative
Technology for the Earth
economies and least developed countries grow at the rate of around 2%/year and around
3%/year even in 2100, respectively. The global average growth rate from 2000 to 2100 is
2.1% per year.
Both in Scenario A and B, economic gaps between developed and developing countries
narrow steadily until 2100. Yet, the gaps in Scenario B are still bigger than in Scenario
A. The GDP per capita ratio of OECD90 to Africa is 38.5 in 2000, and in 2100 6.4 in
Scenario A and 6.7 in Scenario B.
Figure 4 per capita GDP Scenario
GDP scenarios are formulated by combination population scenario and per capita GDP
scenario. Figure 8 shows the world GDP scenarios. The potential world GDP grows at a
higher rate in Scenario B than in Scenario A. As mentioned above, GDP per capita
growth at a higher rate in Scenario B makes the population smaller than the population
in Scenario A, so that GDP difference between the two scenarios shrinks. The world
average of GDP annual growth is assumed to be 2.0% per year in Scenario A and 2.3%
per year in Scenario B from 2000 to 2100.
10
Research Institute of Innovative
Technology for the Earth
Figure 5 GDP (MER) Scenarios
(Reference) Systems Analysis Group, RITE, Development of Long-term Socioeconomic
Scenarios -Population, GDP-, August 15, 2011
http://www.rite.or.jp/English/lab/syslab/research/alps/baselinescenario/E-ScenarioOutli
ne_POPGDP_20110815.pdf
2.2. Demand
Population and GDP are not directly utilized to project future energy system, rather to
assume the level of production or extent of service activity for individual sectors. The
projected level of production or service activity is consistently satisfied by the optimal
combination of various bottom-up technologies for the sectors that are explicitly
modeled. For the other sectors, baseline amounts of final energy demands are assumed
together with their long-term price elasticity using top-down modeling without
explicitly describing bottom-up technologies.
The rest of final energy demands are estimated in a top-down manner, represented by
four type of energy careers, which include solid energy, liquid energy (gasoline, light oil,
and heavy oil), gaseous fuel and electrical energy, are assumed for aggregated three
sectors: industry, transportation and residential and commercial.
2.3. Macro-economy
(see Population and GDP)
11
Research Institute of Innovative
Technology for the Earth
2.4. Technological change
Technological change is generally treated exogenously.
2.5. Behavioural change
DNE21+ is a least-cost optimization model that provides a detailed representation of
energy supply and energy technology, modeling technology choice behavior of investors
or consumers in a bottom up manner. Payback period, operation and maintenance cost,
energy prices, technology costs and performance parameters determine the least-cost
energy-equipment combination that meets a specific energy need in the model.
The payback period is a key parameter in determining their behavior, affected by
numerous kinds of factors observed in the society such as interest rate, the depreciation
rate, the price change rate of capital goods, income, subjective preference for risk and
prospective profit rate of stockholders. A number of study reveals that payback period
varies widely among technologies, countries, and sectors.
In business behavior, the return on investment (ROI) is generally 10–20%, and this
means that payback period has to be 5–10 years. Most of large Japanese companies in
industrial and commercial sectors make investment decisions in energy-saving
technologies with 3–5 years payback periods (Energy Conservation Center, Japan
[ECCJ] 2004). The payback period for the purchase of light-duty vehicles is 1.8–5 years
(US EPA 2005). The payback period of consumer durable goods, such as space-heating
systems, air conditioning, and refrigerators is generally 1-3 years or shorter (Wada et al.
2012; Train 1985; Dubin 1992).
Furthermore, the payback periods in developing countries are shorter than those in
developed countries, and those in the residential and commercial sectors are shorter
than those in industrial sectors.
Table 2 shows the payback periods of DNE21+, which come close to matching the
observed payback periods in the real world, although the observed payback periods in
different countries, sectors, and technologies are limited and uncertain. The model
assumes different payback periods based on economic stages across countries. The
periods become longer in accordance with the growth of economic level.
12
Research Institute of Innovative
Technology for the Earth
The selections of energy technologies and CO2 emissions for 2005 determined within the
model are roughly calibrated with the historical data by adjusting the assumptions of
payback periods.
Table 2 Payback period and implicit discount rate
Payback period (Implicit discount rate)
Upper limit Lower limit
Electricity generation sector 11.9 (8%) 5 (20%)
Other energy conversion sector 6.6 (15%) 4 (25%)
Industrial sector (energy-intensive industry) 6.6 (15%) 4 (25%)
Transportation sector
(Purchase of environment-conscious products)
3.3 (30%) 2.2 (45%)
10 (10%)
Residential & commercial 3.3 (30%) 1.8 (55%)
3. Energy
DNE21+ includes a detailed description of energy carriers and conversion technologies.
It includes eight types of primary energy sources; coal, oil (conventional and
unconventional), natural gas (conventional and unconventional), hydro power and
geothermal, nuclear, wind power, photovoltaics and biomass). Interregional
transportation of energy (coal, oil natural gas, synthetic oil, ethanol, electrical power
and hydrogen) and CO2 are incorporated in the model. As technological options, various
types of energy-conversion technologies are explicitly modeled, such as electricity
generation, oil refinery, natural gas liquefaction, coal gasification, and water
electrolysis, methanol synthesis. The end-use sector are disaggregated into four types of
secondary energy carriers: 1) solid fuel, 2) liquid fuel, 3) gaseous fuel, and 4) electricity.
The demands for these energy carriers are endogenously calculated in a top-down
fashion using long-term price elasticity in other cases than the reference case.
13
Research Institute of Innovative
Technology for the Earth
Figure 6 Outline of energy flows in DNE21+
3.1. Energy resource endowments
3.1.1. Fossil Reserves and Resources
Estimation of fossil fuel reserves refers to a number of studies, and supply cost curves
for each resources are made based on economic and technological assumptions. For coal
resource assessment “Survey of Energy Resources” by the World Energy Council (WEC)
is mainly referred to. The USGS 1995 National Assessment of United States Oil and
Gas Resources, the USGS Survey world petroleum assessment 2000: Description and
results, and Rogner (1997) were used for conventional/non-conventional coal and gas
reserve estimation.
The resource potentials are modeled using region-specific potentials, and they are
classified into different grades. With IEA World Energy Outlook and Rogner (1997),
fossil fuel supply curves are created as shown below. This grade structure with royalty
and transportation costs allows to calculate supply chain optimization of global energy
systems.
Fossil fuelsCoalOil (conventional, unconv.)
Gas (conventional, unconv.)
Cumulative production
Unit
production
cost
Renewable energiesHydro power & geothermalWind power
PhotovoltaicsBiomass
Annual production
Unit
supply
cost
Nuclear power
Energy conv.
processes
(oil refinery, coal gasification, bio-ethanol, gas
reforming, water electrolysis etc.)
Industry
Electric
Power
generation
CCS
Transport
Residential & commercial
Iron & steel
Cement
Paper & pulp
Chemical (ethylene, propylene,
ammonia)
Aluminum
vehicle
Refrigerator, TV, air conditioner
etc.
Solid, liquid and gaseous fuels, and
electricity <Top-down modeling>
Solid, liquid and gaseous fuels, and
electricity <Top-down modeling>
Solid, liquid and gaseous fuels, and
electricity <Top-down modeling>
14
Research Institute of Innovative
Technology for the Earth
Figure 7 Cumulative Global Coal Supply Curve
Figure 8 Distribution of Conventional Oil Resources (Source) USGS
15
Research Institute of Innovative
Technology for the Earth
Figure 9 Cumulative Global Oil Supply Curve (incl. Non-conventional)
Figure 10 Distribution of Conventional Natural Gas Resources (Source) USGS
16
Research Institute of Innovative
Technology for the Earth
Figure 11 Cumulative Global Gas Supply Curve (incl. Non-conventional)
In DNE21+, emissions from fossil fuel combustion can be curbed by deploying carbon
capture and storage (CCS). Storage potential was estimated based on a sedimentary
basin map of USGS. The “ideal” potential of aquifer sequestration is shown in Figure
12.
Figure 12 CO2 Sequestration Potential into Aquifer
3.1.2. Renewable Resources
The resource potentials for solar and wind are estimated by using physical data
combined with global land cover data developed by Chiba University. With globally
gridded wind speed data provided by the National Oceanic and Atmospheric
17
Research Institute of Innovative
Technology for the Earth
Administration (NOAA) National Climatic Data Center (NCDC) and land use/cover GIS
data, wind potentials are estimated as shown in Figure 13. The region-specific
potentials are classified into five grades, and the technical potentials for wind power
amount to 13,750 TWh/yr. Potentials of photovoltaics are estimated by solar radiation
data offered by the National Aeronautics and Space Administration, Sea-viewing Wide
Field-of-view Sensor (SeaWiFS) Project, and land-use data. Figure 15 overlays solar
radiation intensity on a global map. In total, the solar potentials amount to 1,270,000
TWh/yr.
Wind power and photovoltaics is assumed to have an annual costs decrease rate of 1.0%
and 3.4%, respectively. In 2000, the unit costs of wind power is 56– 118$/MWh and
photovoltaics 209–720$/MWh, depending on wind velocity and solar radiation etc. In
2050, the unit costs of wind power and photovoltaics are assumed to become
34–71$/MWh and 37–128$/MWh, respectively.
Figure 13 Resources of wind power
18
Research Institute of Innovative
Technology for the Earth
Figure 14 Regional supply potential for wind power (TWh/yr)
Figure 15 Solar radiation intensity (Annual Average)
19
Research Institute of Innovative
Technology for the Earth
Figure 16 Regional supply potential for solar power (PV) (TWh/yr)
Currently hydropower plays an important part in global power generation and is the
most common form of renewable energy. The overall technical potential for hydropower
is estimated to 25,000 TWh/yr, using the WEC’s “Survey of Energy Resources” as a
reference.
20
Research Institute of Innovative
Technology for the Earth
Figure 17 Regional supply potential of Hydro and Geothermal (TWh/yr)
For biomass resource assessment, DNE21+ employs the LULUCF model results on
available land area for biomass production and afforestation, and land are productivity.
Waste-based biomass potentials are also taken into calculation as a constraint of the
DNE21+ model. Exogenous scenario is given for the future traditional biomass.
21
Research Institute of Innovative
Technology for the Earth
Figure 18 Available land potential for cellulosic biomass or afforestation
Figure 19 Available land potential for Biomass Residues
22
Research Institute of Innovative
Technology for the Earth
Figure 20 Available land potential for Traditional Biomass
3.2. Energy conversion
DNE21+ model covers various type of energy conversion technologies, including
electricity generation, coal gasification and liquefaction, natural gas reforming, and
carbon dioxide capture, storage and sequestration (CCS) for energy conversion process.
3.2.1. Electricity
The modeled electricity generation options include: Coal power {low efficiency
(subcriticality), mid-efficiency (supercriticality), high efficiency (extra
supercriticality–IGCC/IGFC), and IGCC with pre-combustion CO2 capture}, Oil power
{low efficiency (diesel generator, etc.), mid-efficiency (subcriticality), high efficiency
(supercriticality), and CHP}, Synthetic oil power {mid efficiency, and high efficiency},
Natural gas power {low efficiency (steam turbine), mid-efficiency (conventional NGCC),
high efficiency (high temperature NGCC), CHP, and oxy-fuel combustion}, Biomass
power {low efficiency, and high efficiency}, Nuclear power {conventional, and
next-generation (Generation IV, etc.)}, Hydro/geothermal power, Wind power, and
Photovoltaics. In association with generation technologies, Power storage system for
wind/PV, Hydrogen power, Electrical cable {conventional, superconducting high
efficiency}, and CCS {post-combustion capture; applicable for coal, oil, synthetic oil,
natural gas, biomass power} are also represented in DNE21+.
As shown above, each type of power generation technology has is classified according to
23
Research Institute of Innovative
Technology for the Earth
level of energy efficiencies and facilities costs are differentiated corresponding to the
level of efficiency. The different levels of generation efficiencies are assumed in order to
represent the broader ranges in current generation efficiency levels in different
countries (see Oda et al. 2012). Their technological progresses are assumed exogenously.
Table 3 shows the assumptions on capital costs and the efficiency of electricity
generation. Fossil fuel prices are endogenously determined within the model by using
the relationship between the cumulative production of fossil fuels and production costs.
However, the fossil fuel prices will be dominated not only by production prices, but also
by speculation, etc. Therefore, the baseline fossil fuel prices are calibrated to meet the
prices of the reference scenarios of the IEA WEO 2010 (IEA 2010b) over the assessment
time periods, while the prices in the mitigation scenarios are endogenously determined
by the cumulative amounts of production induced by levels of emission reductions or the
MAC. DNE21+ also tracks investments by vintage capital stock.
24
Research Institute of Innovative
Technology for the Earth
Table 3 Capital costs and generation efficiency
*Some of capital costs and efficiency are shown in a range because they change over time.
Electricity demand is modeled in a way that demand-supply is balanced. The demand is
expressed by the load duration curves, representing four time periods, instantaneous
peak, peak, intermediate, and off peak time periods, in accordance with the level of
electricity demand. This enables appropriate evaluation of electricity system
corresponding to the characteristics of individual power generation technologies such as
the base power load power plants and the peak load power plants.
For nuclear power generation, exogenous scenarios are assumed for nuclear power
generation up to 2030. Some constraints are assumed that the power generation of
nuclear would be capped at 50% of the total power generation amount and that an
annual expansion of conventional nuclear power generation would be 0.33%, and the
[
文
書
の
重
要
な
部
分
を
引
用
し
て
読
者
の
注
意
を
引
い
た
り
25
Research Institute of Innovative
Technology for the Earth
expansion rate of advanced nuclear power generation would be 1%. As long as the
constraints are obeyed, costs-efficient options are selected by the model.
(Source) Akimoto, Keigo, et al. "Comparison of marginal abatement cost curves for
2020 and 2030: longer perspectives for effective global GHG emission reductions."
Sustainability Science 7.2 (2012): 157-168.
3.2.2. Other conversion processes
Besides electricity and heat generation there are three further subsectors of the
conversion sector represented in DNE21+, such as oil refinery, natural gas liquefaction,
coal gasification, and water electrolysis, methanol synthesis.
3.2.3. Grid and infrastructure
Inter-regional energy transmission infrastructure, such as pipelines for liquid and gas,
such as oil, natural gas, synthetic oil, ethanol, hydrogen and CO2, and power grids, are
represented in the DNE21+ model.
In terms of systems integration, wind power and solar PV are represented in the
DNE21+ model as follows:
(1) Capacity credit:
There are some literatures that evaluate capacity credits of wind power in the United
States and Europe (e.g., Milligan and Poter 2008, Holttinen et al. 2009). The estimated
capacity credits of wind power vary widely from approximately a few percent to 40% by
region. It is also observed that there is a correlation between the capacity credit and the
level of technology penetration: the capacity credit becomes lower in higher wind power
penetration. When the share of wind power capacity in peak load is 30%, the capacity
credits of wind power range from 5% to 25% (Holttinen et al. 2009). In addition, the
methods used for the evaluation of the capacity credit exist widely by region, such as
capacity factor in peak period and equivalent load carrying capacity.
For solar PV, GE Energy (2010) reported that the capacity credit of solar PV is higher
than that of wind power according to the study by the WestConnect group in the United
States. In Japan, the capacity credit of solar PV in summer is considered as 16%
(Japanese government committee on electricity supply and demand 2013). However,
available studies that evaluate the capacity credit of solar PV are limited compared
with wind power.
26
Research Institute of Innovative
Technology for the Earth
In the DNE21+ model, capacity credit is defined as potential power supplies from wind
power and solar PV without electricity storage at the instantaneous peak. Since the
peak of these generation does not always match the instantaneous peak time period of
power demand, the output of wind power generation at instantaneous peak time is
constrained in the model. The capacity credit of wind power is assumed to be 10% in all
regions. Although the physical situation for solar and wind energy is different, the same
assumption with wind power is applied to solar PV in this paper.
(2) Grid stability
Capacities of wind power and solar PV without electricity storage are limited for the
grid stability. In DNE21+, maximum shares in the total electricity supply are 10% both
for wind power and PV without electricity storage. Electricity storage systems on the
demand side are required for wind power and solar PV to be installed over that shares.
If wind power and solar PV are deployed with electricity storage, further 20% of the
total electricity supply are available from wind power and solar PV as additional
capacities. The capital cost of electricity storage is exogenously assumed to be
1600$/kWh (2005) – 40$/kWh (2050), presuming rapid technology progress for
electricity storage.
Theoretically, the maximum share of wind power and solar PV together in the total
electricity generation reaches 60% (10% for wind power without storage, 20% for wind
power with storage, 10% for solar PV without storage and 20% for solar PV with
storage). The recent large regional wind integration studies in the United States
(Milligan et al. 2009) have evaluated wind energy generates up to 30% of annual energy
demand. The outlook of electricity generation shares of wind power and solar PV is16%
and 20% in 2020 and 2030, respectively, in EU according to the EC communication (EC
2010). The assumed total maximum share is suitable level for energy system
assessment until 2050 considering these targets.
The water electrolysis for hydrogen production by photovoltaics has no upper limit,
(naturally restrictions on supply of natural resources should be treated separately).
(Reference) Impacts of different diffusion scenarios for mitigation technology options
and of model representations regarding renewables intermittency on evaluations of CO2
emissions reductions, Fuminori Sano, Keigo Akimoto, Kenichi Wada
27
Research Institute of Innovative
Technology for the Earth
3.3. Transport
DNE21+ represents road transportation sector in two way. One is vehicle type and the
other is technological category. The type of vehicle includes small passenger car, large
passenger car, bus, small truck, and large truck. Vehicle technologies are categorized
into internal combustion engines, electric cars, fuel-cell cars, and alternative fuel
vehicles, including bioethanol mixed with gasoline, biodiesel mixed with diesel, and
CNG. The gasoline and diesel combustion engines for gasoline/diesel are further
classified into conventional internal combustion cars (low/high efficiency), hybrid cars,
and plug-in hybrid cars.
Figure 21 Schematic diagram of transportation service demand in DNE21+
Gasoline
Ethanol
Passenger car
(Small)
Passenger car
(Large)
Bus
Truck
(Small)
Truck
(Large)
Passe
nger tra
nsp
ort [p
-km]
Fre
ight tra
nsp
ort [t-km
]
Gasoline Engine base
Diesel Engine base
Electricity
Hydrogen
Light Oil
Gasoline +
Ethanol (<20%)
Gasoline +
Ethanol (≧20%)
Bio diesel
Light oil +
Bio diesel (<20%)
Light oil +
Bio diesel (≧20%)
Passenger car
(Small/Large)
Bus
Truck
(Small/Large)
Passenger car
(Small/Large)
Bus
Truck
(Small/Large)
Passenger car
(Small/Large)
Bus
Truck
(Small/Large)
Passenger car
(Small)
Truck
(Small)CNG
28
Research Institute of Innovative
Technology for the Earth
Figure 22 Taxonomy of passenger cars in DNE21+
Scenarios on service demand of road transportations are developed for passenger cars
and buses separately based on per-capita GDP and the historical trends. As for road
freight transport (t-km) scenarios of cargo trucks, overall cargo service per-capita is
estimated by the GDP size, and then the transition of modal share is assumed.
Figure 23 Traffic service of passenger car by region
ICEV:従来型内燃機関車、HEV:ハイブリッド車、HEV(Plug-In):プラグインハイブリッド車、EV:純電気自動車、FCHEV:燃料電池車、B:高濃度バイオ燃料利用車
EV
ICEV-Diesel
(Low-Eff.)
ICEV-Diesel
(High-Eff.)
HEV-Diesel
HEV-Diesel
(Plug-In)
ICEV-B-Diesel
HEV-B-Diesel
HEV-B-Diesel
(Plug-In)
Fuel :
Gasoline
Ethanol
Light oil
Bio diesel
Gasoline +
Ethanol (<20%)
Gasoline +
Ethanol (≧20%)
Light oil +
Bio diesel
(<20%)
Light oil +
Bio diesel
(≧20%)
ICEV-Gasoline
(Low-Eff.)
ICEV-Gasoline
(High-Eff.)
HEV-Gasoline
HEV-Gasoline
(Plug-In)
ICEV-B-Gasoline
ICEV-CNG
HEV-B-Gasoline/
HEV-CNG
HEV-B-Gasoline
(Plug-In)
Traffic Service
Diesel engine base
Gasoline engine base
FCHEV
Electricity
Hydrogen
CNG
29
Research Institute of Innovative
Technology for the Earth
Figure 24 Traffic service of cargo truck by region
3.4. Residential and commercial sectors
DNE21+ has detailed technology representation for residential and commercial sectors.
The modeled technologies include, refrigerator, lighting, television, air conditioner, and
gas cooking stove. Each of above technology are segmented into several product
category according to size and efficiency as shown in Table 4.The other technologies in
the residential and commercial sectors are aggregated and modeled in a top-down
manner.
30
Research Institute of Innovative
Technology for the Earth
Table 4 Product category of appliances
3.5. Industrial Sector
DNE21+ explicitly models technologies for industry subsector, such as iron and steel,
cement, pulp and paper, aluminum, petrochemical and ammonia. The other subsector
are aggregated and modeled in a top-down manner. Each subsector includes following
technology options:
Iron and steel; Blast Furnace (BF) - Basic Oxygen Furnace (BOF) {low efficiency (small
scale), mid-efficiency (large scale), high efficiency (large scale, equipped with Coke Dry
Quenching (CDQ), Top-pressure Recovery Turbine (TRT), recovery of by-product gases),
next-generation (super coke oven, eg. SCOPE 21, utilizing plastic wastes and tire
wastes, as well as highly efficient equipments), iron making by hydrogen reduction},
coke oven gas (COG) recovery {externally attachable to low/mid-efficient BF-BOF}, basic
oxygen furnace gas (LDG) recovery, CDQ/TRT {externally attachable to mid-efficient
BF-BOF}, Direct reduction {natural gas base (mid/high efficiency), hydrogen
gasification base}, Scrap- Electric Arc Furnace (EAF) {low efficiency (small scale),
mid-efficiency (tri-phase electric arc furnace), high efficiency (DC water-cooled walls arc
31
Research Institute of Innovative
Technology for the Earth
furnace equipped with scrap preheating)}, CCS {applicable for BF-BOF}
Cement; [Small scale facilities] Vertical kiln, Wet rotary kiln, Dry rotary kiln, SP/NSP
dry rotary kiln {equipped with suspensionpreheaters (SP), or new SP (NSP) meaning
precalciner}, Advanced fluidized bed shaft furnace {equipped with SP/NSP, efficient
clinker coolers}, [Large scale facilities (more efficient than small scale)] Wet-process
rotary kiln, Dry-process rotary kiln, SP/NSP dry-process rotary kiln, SP/NSP
dry-process rotary kiln (BAT) {equipped with efficient clinker coolers, SP with 5 or 6
levels, efficient waste heat recovery}
Pulp and Paper; Chemical pulp {low efficiency, mid-efficiency, high efficiency,
next-generation}, Paper recycling {low efficiency, mid-efficiency, high efficiency}, Milling
paper {low efficiency, mid-efficiency, high efficiency, Next-generation}, Black liquid
recovery&use {low efficiency, high efficiency}, Paper sludge boilers, Steam turbine
power systems
Aluminum; Söderberg aluminum production, Prebake aluminum production
Chemical; Ethylene/propylene: Naptha cracking {low efficiency, mid-efficiency, high
efficiency, next-generation}, Other production {ethane cracker etc. low efficiency,
mid-efficiency, high efficiency}
Ammonia: from Coal {low efficiency, mid-efficiency, high efficiency}, from Oil {low
efficiency, mid-efficiency, high efficiency}, from Natural gas {low efficiency,
mid-efficiency, high efficiency}
As an illustration how these technological options are model in DNE21+, the outline of
the modeling framework for the iron and steel sector is shown below:
1. Nine types of steelmaking routes having different levels of energy efficiency are
modeled. These routes include four types of BOF steelmaking, three types of
scrap-based EAF steelmaking and two types of DRI-based EAF steelmaking.
2. In the BOF steelmaking routes, retrofit measures of the facilities for CDQ, TRT,
waste plastics and tires recycling, and COG and LDG recovery are explicitly
modeled.
3. The lifetime of all the facilities in the steel sector described above is assumed to be
40 years. The model considers the historical installation of the facilities.
4. Scenarios of crude steel production by region are assumed exogenously. In addition,
32
Research Institute of Innovative
Technology for the Earth
the maximum and minimum scrap-based EAF steel production scenarios are also
assumed. These assumptions are kept fixed regardless of the simulation cases.
Figure 25 shows the concept of energy flows in steelmaking process modelling. These
nine steelmaking routes encompass processes from raw materials input to coke oven
and sintering furnace and from scrap input to EAF and BOF, to hot rolling. The
processes of downstream, such as cold rolling, thin coating, special steel making, and
ferroalloy making are not considered.
Figure 25 Modeling of energy use of the steel sector in DNE21+.
As shown in Figure 26, DNE21+ assumed that the low-efficiency basic oxygen furnace
(BOF) steelmaking route (type I) has a smaller scale capacity, partly including ingot
making and some classical processes such as beehive coke oven and open hearth furnace
(OHF). Type I is allowed to retrofit coke oven gas (COG) recovery in the model.
33
Research Institute of Innovative
Technology for the Earth
Figure 26 Type I energy use of low-efficiency BOF steelmaking (
Fig. 19 shows the Type II energy use for middle-efficiency BOF steelmaking route. Type
II is a large-scale facility with modern steelmaking processes including pulverized coal
injection (PCI) and continuous casting facilities. The average coal injection ratio in the
type II is 88 kg/t-pig iron (2.3 GJ/t-pig iron), which can bring a net energy saving of
1.0–1.4 GJ/t-pig iron. The model allows some retrofit measures for COG recovery, basic
oxygen furnace gas (LDG) recovery, effective utilization facility of COG and LDG, CDQ,
and TRT to the type II.
Figure 27 Type II energy use of middle-efficiency BOF steelmaking
34
Research Institute of Innovative
Technology for the Earth
The high-efficiency BOF steelmaking route (type III) shown in Figure 28 is for a
large-scale facility with sophisticated steelmaking processes including CDQ, TRT and
COG, BFG, and LDG recovery and effective utilization facilities of the gases and
sensible heat. The average coal injection ratio in the type III is 151 kg/t-pig iron (4.0
GJ/t-pig iron), which can bring a net energy saving of 1.5 GJ/t-pig iron. Energy input
data in the figures takes into consideration the net energy saving effects derived from
these energy efficient facilities. Type IV is the steelmaking process with the
next-generation coke oven added to type III. Type IV is assumed to be available only
after 2011. All the steelmaking routes (type I–IV) are assumed to use scrap of 161
kg-scrap/t-crude steel in BOFs.
Figure 28 Type III and IV energy use of high-efficiency BOF steelmaking
Figure 29 shows the energy flow for the electric arc furnace (EAF) steel production. The
low efficient electric steelmaking route (type V) consists of a small-scale EAF and
induction furnace that is widely utilized in India. The mid-efficient EAF steelmaking
route (type VI) assumes an alternate current (AC) arc furnace that is widely used in the
US, Europe, and Japan. The highly efficient EAF steelmaking route (type VII) consists
of a direct current (DC) arc furnace and many types of energy saving facilities such as
scrap preheating and recuperative burner ladle preheating.
35
Research Institute of Innovative
Technology for the Earth
Figure 29 Modeling of three types of EAF steelmaking (type V, VI, and VII).
The correlation between evolution of per-capita GDP and per-capita apparent
consumption of crude steel, trends in industry structure change by region, government
planning reports etc. were taken into account in the scenario shown in Figure 30. The
crude steel production is modeled by sorting out the processes into three routes; basic
oxygen furnace (BF-BOF), scrap-based electric arc furnace (EAF) and DRI-based
electric arc furnace (DRI-EAF) where the estimation on the historical installation of the
facilities are conducted and their results are taken into account; installation year,
energy efficiency and capacity.
36
Research Institute of Innovative
Technology for the Earth
Figure 30 Crude-steel production of major regions (statistics and future scenario)
4. Land Use
Potentially available land area for energy crop production and afforestation, which is
one of the input data of DNE21+ model, is evaluated by using a land-use model to
maintain consistency with the land area required for food crop production and forest
conservation and water-stressed basins. The land-use mode is basically a
15-minute-grid model, for every decade from 2000 to 2050, and at time points for 2070
and 2100. It is integrated with the water-use model; therefore, the water-stressed
basins are estimated under the same scenario regarding socioeconomic development,
climate change, and land use for food crop production.
a) Land area required for food crop production
Food crop types include wheat, rice, maize, sugar cane, soybeans, oil palm fruit,
rapeseed and others, in consideration of their importance in terms of principal food,
favorite food, vegetable oil and feed considerations. The available land area for food crop
production are allocated in order to meet the regional food demands. If there is a
shortage of food crop production due to a lack of available land, production is reallocated
to regions where land is still available and any shortage is fulfilled by means of trade.
OECD Europe
OECD North America
OECD Pacific
China
India
Other developing Asia
Economies in transition
Africa and Middle East
Latin America
0
500
1,000
1,500
2,000
1990 2000 2010 2020 2030 2040 2050
Cru
de
ste
el p
rod
uct
ion
(M
illio
n t
on
s o
f cr
ud
e s
tee
l/yr
)
Statistics
Scenario
37
Research Institute of Innovative
Technology for the Earth
Impacts due to climate change and adaptations for planting can be taken into account
through changes in the potential production, which can be calculated using the
information on crop characteristics and soil types provided in the in the Global
Agro-Ecological Zones (GAEZ) model (Fischer et al. 2002) for both rain-fed and irrigated
conditions. Impacts on crop productivity due to technological progress associated with
economic growth and production adjustments to avoid falling crop prices and so on are
taken into account through a scenario of ‘yield (or management) factor’ for each of crops
and for each of the 32 regions. For all crops, a greater growth of the ‘yield factor’ is
expected in developing regions than in developed regions. The grids available for
irrigation are based on an irrigation map for the year 2000 (Siebert et al. 2007). The
regional areas for irrigation maintain the consistency with those of the original map,
although the share of irrigation area for each of grids is simplified to be either 0 or 100%.
No expansion of the irrigation available grids is allowed in the future according to the
assumptions by Alcamo et al. (2007). For details of the land-use model, please refer to
Hayashi et al. (2013).
(Reference) Hayashi, Ayami, et al. "Global evaluation of the effects of agriculture and
water management adaptations on the water-stressed population." Mitigation and
Adaptation Strategies for Global Change (2013): 1-28.
b) Available land area for energy crop production and afforestation
The agro-land use model is equipped with a land cover map for eight major land-cover
categories (i.e., rainforest, other forests, arable land, grassland, pasture, barren,
built-up, and water), which was constructed by reference to data from around 2000
(Fischer et al. 2008, PBL 2009, USGS 2000). Fallow land is estimated by subtracting the
land required for food crop production from arable land. Then, fallow land and
grassland are treated as candidate land for energy crop production and afforestation.
They don’t include either cropland or pasture; therefore, there is no worrying about
competition with food production technically. Furthermore, they don’t include
rainforests and other forests; therefore, concerns regarding CO2 emissions and
biodiversity loss through the use of land for energy crop production are minimal. In the
next step, water-stressed basins are excluded from the candidate land. Water-stressed
basins are estimated based on a criterion (0.4 ≤ the annual water
withdrawal-to-availability ratio) while maintaining consistency in terms of agricultural
water use for irrigation, domestic and industrial water use, and water availability. Data
38
Research Institute of Innovative
Technology for the Earth
for the river basins are derived from the Total Runoff Integrating Pathways (TRIP)
database (Oki 2001).
We focus on land satisfying conditions for energy crop yield and land accessibility of the
candidate land, and we calculate the available energy from the energy crop production
on the land by level of conditions for yield and land accessibility. Energy crop types
include cereals (wheat, rice, or maize), sugar cane, oil palm fruit, other oil crops
(soybeans, or rapeseed), and lignocellulosic crops.
c) CO2 emissions from LULUCF
Four types of CO2 emission/fixation are taken into account; (1) CO2 emissions due to
expansions of crop land for food crop production, (2) CO2 fixation by forests planted and
naturally-regenerated before 2010, (3) CO2 fixation by afforestation after 2010, and (4)
CO2 emissions by other sources (i.e., emission abandoned land after deforestation).
CO2 emission due to expansions of crop land for food crop production is estimated by
multiplying the land area converted to crop land from other types of land, by CO2
emission coefficients. The former is evaluated by using the land-use model, and the
latter was constructed by reference to Houghton’s studies (1999, 2001) for each of the 32
regions.
Land area for forests planted and naturally-regenerated during the last 60 years were
estimated based on the FRA 2010 report (FAO, 2010) for each of the 54 regions. The
amount of CO2 fixation by the forests were calculated based on the estimated forest area
and the NPP (Net Primary Production) for each of the regions. CO2 fixation by
afforestation after 2010 is one of the mitigation options in the DNE21+ model.
CO2 emissions by the other sources in 2005 were estimated so that the total CO2
emission to consists to the amount for the year by RCP database (IIASA database). After
that, it was assumed to decrease associated with increase in per capita GDP.
5. Non-energy CO2 emissions
CO2 emissions from industrial processes, such as cement production, are accounted for
based on the cement production scenario.
39
Research Institute of Innovative
Technology for the Earth
6. Non-CO2 GHGs
The non-CO2 GHG assessment model, one of the DNE21+ model group, disaggregates
the world into the 54 regions, consistent with the regional definition in the DNE21+
model (Akimoto et al.(2009), Akimoto et al.(2009)). The model converts bottom-up
assessments of mitigation technologies performed by the USEPA (2002) to a proxy
model using elasticity (Dowaki et al., 2006). The historical emission of non-CO2 GHGs
was adjusted using the UNFCCC inventory (UNFCCC, 2009) and the IEA statistics
(IEA, 2007) for Annex I countries and non-Annex I countries, respectively. The emission
mitigation costs and potentials were modified using new technology assessments
performed by the USEPA (2006b).
Equation (1) indicates the relationship between the individual non-CO2 GHG mitigation
ratio and marginal abatement costs evaluated by using the elasticities. In the model,
the non-CO2 GHG mitigation in 54 regions is estimated when the non-CO2 GHG
abatement costs are equalized to the CO2 marginal abatement costs. The elasticities are
determined such that the marginal abatement cost curves correspond to the USEPA
estimates obtained separately for each sector and type of gas. The estimates are
calculated using a technology database for non-CO2 GHG measures. Thus, the model
used here is not a direct bottom-up model; however, marginal costs and potentials of
non-CO2 GHG mitigation are essentially based on the bottom-up analysis of the
USEPA.
Basically the elasticities in the model developed by Hyman et al. (2003) are applied to
the model, but the elasticities are adjusted to be consistent with the result of analysis
for a 20%/year discount rate which are close to the rate usually observed in decision
makings of socio-economic activities, taking into consideration the results of sensitivity
analysis for the discount rate (payback period) carried out by the USEPA (2002). The
elasticities are also partially adjusted on the basis of the mitigation effect report of the
USEPA (2006b). The elasticities of gases and the mitigation potentials in the model are
estimated to be less than those reported by Hyman et al (2003).
n)h,σ(g,
t)n,h,P(g,
11t)n,h,Red(g,
(1)
where g represents the gas; h, the sector; n, the region; and t, the year. Red is the
reduction rate for the total emission, P is the marginal abatement cost, and σ is the
elasticity derived on the basis of studies by the USEPA (2002, 2006b) and Hyman et al.
(2003).
40
Research Institute of Innovative
Technology for the Earth
The baseline emissions were estimated using the above assumptions for population and
GDP adopted in the model. The model considers five types of emissions: CH4 in seven
sectors, N2O in six sectors, hydrofluorocarbons (HFCs) in one sector, perfluorocarbons
(PFCs) in one sector, and SF6 in one sector for 54 regions.
CH4 emissions were considered in seven sectors: agriculture, oil, coal, natural gas,
residential, transportation, and energy intensive industries. N2O emissions were
considered in six sectors: agriculture, oil, natural gas, residential, transportation, and
energy-intensive industries. HFCs, PFCs, and SF6 emissions were considered for one
macro-sector each. The baseline HFCs, PFCs, and SF6 emissions were basically
estimated using the SRES B2 scenario for each of the four regions the world was divided
into.
(Reference)
Akimoto, Keigo, et al. "Estimates of GHG emission reduction potential by country, sector,
and cost." Energy Policy 38.7 (2010): 3384-3393.
Dowaki, K., K. Akimoto, F. Sano, T. Tomoda, S. Mori, 2006. An Impact Analysis on
Greenhouse Gases Including an Effect of Non-CO2 Emissions, Operations Research,
Germany.
Hyman, R.C., J.M. Reilly, M.H. Babiker, A. Valpergue De Masin, H.D. Jacoby, 2003.
Modeling non-CO2 greenhouse gas abatement, Environmental Modeling and
Assessment, Vol. 8, 175-186.
USEPA, 2002. International Methane and Nitrous Oxide Emissions and Mitigation
Data, http:// www.epa.gov/methane/appendices.html (accessed in February 2009).
USEPA, 2006a. Global Anthropogenic Non-CO2 Greenhouse Gas Emissions: 1990-2020.
USEPA, 2006b. Global Mitigation of Non-CO2 Greenhouse Gases, http://www.epa.gov/
climatechange/economics/downloads/GlobalMitigationFullReport.pdf. (accessed in
February 2009).